Casadesus Masanell, Zhu - Business Model Innovation - R2 PDF

Business Model Innovation and Competitive Imitation: The Case of
Sponsor-Based Business Models
The Harvard community has made this article openly available.

Please share how this access benefits you. Your story matters.
Citation
Casadesus-Masanell, Ramon, and Feng Zhu. "Business Model

Innovation and Competitive Imitation: The Case of Sponsor-Based
Business Models." Strategic Management Journal 34, no. 4 (April
2013): 464482.
Published Version
doi:10.1002/smj.2022
Accessed
August 13, 2016 7:34:21 PM EDT
Citable Link
http://nrs.harvard.edu/urn-3:HUL.InstRepos:13212599
Terms of Use
This article was downloaded from Harvard University's DASH

repository, and is made available under the terms and conditions
applicable to Open Access Policy Articles, as set forth at
http://nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-ofuse#OAP
(Article begins on next page)
Business Model Innovation and Competitive Imitation:

The Case of Sponsor-Based Business Models
Ramon Casadesus-Masanell
Professor
Harvard Business School
Boston, MA 02163
Phone: +1 (617) 496-0176
Email: casadesus@gmail.com
Feng Zhu
Assistant Professor
Marshall School of Business
University of Southern California
Los Angeles, CA 90089-0808
Phone: +1 (213) 740-8469
Email: fzhu@marshall.usc.edu
October 31, 2011
We thank two anonymous referees and Constance Helfat for helpful comments that have resulted in a
much better paper. We also thank Paul Adler, Ana Leticia Gonzalez-Palomares, Joan Enric Ricart, and
Monic Sun for helpful discussions. Casadesus-Masanell thanks the HBS Division of Research.
Business Model Innovation and Competitive Imitation:

The Case of Sponsor-Based Business Models
Abstract
We study sponsor-based business model innovations where a firm monetizes its product
through sponsors rather than setting prices to its customer base. We analyze strategic
interactions between an innovative entrant and an incumbent where the incumbent
may imitate the entrants business model innovation once it is revealed. The results
suggest that an entrant needs to strategically choose whether to reveal its innovation
by competing through the new business model, or conceal it by adopting a traditional
business model. We also show that the value of business model innovation may be
so substantial that an incumbent may prefer to compete in a duopoly rather than to
remain a monopolist.
Key words: business model innovation; imitation; sponsor-based business model; strategic revelation; strategic concealment
Business Model Innovation and Competitive Imitation
INTRODUCTION
Schumpeter (1934) distinguishes between five types of innovations: new products, new methods of production, new sources of supply, exploitation of new markets, and new ways to organize business. Much of the literature so far has focused on the first two types of innovation
(e.g., Shan, Walker, and Kogut 1994; Banbury and Mitchell 1995; Eisenhardt and Tabrizi
1995; Schroeder 2006; Katila and Chen 2008; Leiblein and Madsen 2009; Roberts 1999; Adner
and Kapoor 2010; Leiponen and Helfat 2010; Zhou and Wu 2010). Our study focuses on the
last type of innovation, often referred to as business model innovation today. Business model
innovation has become increasingly important both in academic literature and in practice
given the increasing number of opportunities for business model configurations enabled by
technological progress, new customer preferences, and deregulation.
At root, business model innovation refers to the search for new logics of the firm, new ways
to create and capture value for its stakeholders, and focuses primarily on finding new ways to
generate revenues and define value propositions for customers, suppliers, and partners (e.g.,
Amit and Zott 2001; Magretta 2002; Zott and Amit 2007, 2008; Baden-Fuller et al. 2008;
Casadesus-Masanell and Ricart 2010; Gambardella and McGahan 2010; Teece 2010). As a
result, business model innovation often affects the whole enterprise (Amit and Zott 2001).
New entrants in a wide array of industries have demonstrated time and again that innovative business models can provide the basis for sustainable business success, even in
competitive settings with well-established incumbents. But just as product and process innovations are hard to protect, business model innovations can be imitated: British Airways
(BA) launched Go, a copycat of Ryanairs no-frills model, to compete against European
low-cost airlines; Recoletos, one of the largest Spanish media groups, launched Que!an adsponsored free newspaperin 2005 to fight the entry of similar titles such as Metro Spain;
and in 2007, CBS Interactive copied Hulus media streaming business model.
These empirical observations suggest that incumbents often learn about new business
models from entrants and respond by incorporating these innovations (in full or in part)
into their own businesses. The possibility of competitive imitation, in turn, suggests that
entrants need to strategically choose whether to reveal their ideas by competing through the
new business model or, instead, to conceal them by adopting a traditional, established logic
of value creation and value capture.
While a few theoretical studies have created frameworks to examine competitive dynamics among firms employing different business models (e.g., Lin, Ke, and Whinston 2008;
Casadesus-Masanell and Zhu 2010), these frameworks do not capture the role of innovation
and competitive imitation in firms choices of business models. In this paper, we examine the
desirability or lack thereof of business model innovations when such innovations cannot be
protected and, thus, competitive imitation is possible. Specifically, we ask: under what circumstances will an entrant benefit from adopting a new business model when the innovation
may be imitated by an incumbent?
Given the diversity of business models currently employed by companies in all sorts
of industries, we must constrain the scope of our undertaking by studyingfrom among
the many that exist, could existan important class of business model innovations. We
focus here on business model innovations that allow a firm to monetize its product through
sponsors rather than by setting prices directly to its customer base. We refer to this class of
innovations as sponsor-based business model innovations. To illustrate, consider the following
examples:
Publishing houses traditionally earn their revenues by selling books to readers at positive prices. Alternatively, the publisher could include ads intertwined with the books
text and monetize this content by charging advertisers. In the extreme, the publishing
house could give the books away for free and make money through ads only. Implementing such a scheme would be relatively easy for ebooks, as the ads could change
over time, just as they do in news websites.
The traditional way for porn websites to monetize their content is by charging prices to
surfers. However, some sites are currently competing with a business model whereby
surfers can obtain free porn if they help solve a few captchas to create new, legitimate
email accounts.1 The free email accounts are often worth more to spammers than the
bandwidth porn surfers consume downloading videos and images, and the porn site
can monetize them by selling these email accounts to spammers.
The traditional way ski resorts monetize their offering is by charging skiers positive
prices for access to the slopes. With the growth of timeshare apartments close to the
slopes,2 ski resorts are increasingly partnering with real estate firms in a business model
where skiers are offered free access to the slopes in exchange for enduring several hours
condominium timeshare sales-pitches. In this new scheme, the ski resort monetizes
access to the slopes through revenues obtained from the real estate company instead
of prices paid by the skiers.
When commercial email service appeared in the early 1990s, ISPs such as CompuServe,
Prodigy, and America Online supplied email to paying subscribers through a usagebased billing system and, later, through monthly subscriptions. Launched on July 4,
1996, Hotmail (originally HoTMaiL) was the first free email service. A few months
after launch, Hotmail began displaying advertisements, thus becoming the first adsponsored email service. Implementing such business model required Hotmail to provide access to email through an interface where ads could be easily updated. Hotmail
stored emails on the cloud rather than on users own desktops and most users accessed their accounts through a browser where ads could be updated quickly.
These four examples are instances of the general class of situations that we study. Specifically, there is a traditional business model that involves monetizing the product through
prices charged to consumers (for books; for porn; for access to the slopes; for email ser1
Captchas are scrambled text boxes that many websites use to block bots. For example, Gmail uses
captchas on its account creation page to prevent automatized creation of Gmail accounts which would later
be used to send spam.
2
Timeshare is a form of ownership of real estate (condominiums generally) whereby multiple parties hold
rights to use the property, and each sharer is allotted a period of time (typically one week, and almost always
the same time every year) in which they may use the property.
vice), but an innovator has found a new way to monetize the offering by giving it away to
customers and obtaining revenue from some sponsors. To persuade sponsors to pay, a firm
needs its consumers to provide something to the sponsors in return. Consumers often derive
less utility from the firms product or service as a result: indeed, in the above examples the
customer suffers a reduction in the quality of the good or service, and thus an impoverished
consumption experience (book readers and email users must be exposed to advertisements;
porn surfers must solve some captchas; skiers must sit and listen about timeshare for a
number of hours).
We choose to focus our study on sponsor-based business model innovations for three reasons. First, such innovations appear to be increasingly prevalent in todays economy. For
example, many companies choose to finance themselves using ad revenues and offer their
products or services free to consumers. Such products and services today range from newspapers to software applications, from television programs to online search engines. The
increasing popularity of sponsor-based business models has been partly fueled by opportunities granted by Internet technologies which allow much improved targeting of advertisements
and promotions as well as improved opportunities for direct interaction between firms and
consumers.
Second, while many business model innovations that are to be adopted require full reconfigurations to a firms activity system, sponsor-based business model innovations are
generally not overly burdensome. For example, it is easier for the New York Times to offer
a free, ad-sponsored newspaper than for Ethan Allen to operate like IKEA or for Avis to
reinvent itself into a Zipcar. The implication is that sponsor-based business model innovations seem particularly easy to imitate. Since the purpose of our study is to analyze the
effects of potential imitation on a firms incentives for business model innovation, the case
of sponsor-based business models is most relevant.
Third, although the notion of sponsor-based business model is well-known in sectors of
the media industry, its penetration into other arenas has been a gradual process, as success-
ful implementation of this business model type is not always obvious. For example, while
the modern magazine industry originated in the mid-17th century, magazines began using
advertisements as a means of financial support only late the 19th century.3 Similarly, European no-frills air service providers (such as Ryanair) earn a large share of their revenues
not from ticket prices but from such ancillary sources as subsidies from secondary airports
or payments from bus companies taking passengers from those airports to city centers. Obviously, such revenue sources had been available to traditional flag carriers before the entry
of low-cost airlines. Without a doubt, sponsor-based business models will become feasible in
an ever-increasing number of industries.
Setup and main results
Our study offers the first formal model of business model innovation in a game-theoretic
framework. We focus on sponsor-based business model innovation and provide a comprehensive analysis of strategic interactions between an innovative entrant and an incumbent
where their choices of business models are endogenously determined, and where the incumbent may imitate an entrants innovation once it is revealed.
We analyze a three-stage game with two firms, an entrant and an incumbent, offering
vertically differentiated products. In the first stage, the entrant chooses the business model
through which it intends to compete: it can either compete through the traditional business
model (charging a price to customers for the product) or innovate by adopting a sponsorbased business model (charging zero price and monetizing the customer base in ways that
affect product quality negativelyas noted above). Prior to entry, the incumbent operates
the traditional model and is unaware of the innovation. For simplicity, we assume that firms
face no capacity constraints.
In the second stage, the incumbent observes the entrants business model and chooses its
own in response. If the entrant chooses the traditional business model, then the incumbent
3
Source: http://www.answers.com/topic/magazine, accessed May 2010.
does not learn that there is an alternative way to compete and it may only respond with a
version of the traditional business model (for example, it may introduce several products all
based on the traditional model). However, if the entrant chooses the new business model,
the incumbent can learn the innovation and may choose to imitate it, in full or in part.
In the third stage, firms make their tactical choices about how they will compete within
their choices of business model. We use the expression monetization intensity to refer to
the size of the cost imposed to customers in exchange for the free product. In the examples
above, monetization intensity corresponds to the number of ads in the book or on the website,
to the number of captchas to be solved for access to free porn, and to the number of hours
that skiers must listen to promoters of timeshare condominiums. Obviously, the stronger is
the monetization intensity, the larger is the revenue per customer that the firm derives from
the sponsors. However, as monetization intensity grows, product quality (and consumers
willingness to adopt) deteriorates. In this stage, the entrant chooses price if it entered
with the traditional business model, or monetization intensity if it entered with the new
business model. The incumbent chooses price and/or monetization intensity, depending on
its business model and the entrants choice.
Our analysis provides several new results. First, we find that the entrant will sometimes choose to strategically reveal or conceal its innovation. Strategic revelation refers to
a situation where the entrant prefers to compete through the new business model when it
would choose not to do so if the incumbent was expected to continue competing through
the traditional model. Such revelation induces the incumbent to change its business model
in a way that is beneficial to the entrant. Strategic concealment refers to a situation where
the entrant prefers not to compete through the new business model when it would choose to
do so if the incumbent was expected to continue competing through the traditional model.
Such concealment prevents the incumbent from changing its business model in a way that is
detrimental to the entrant.
We find that strategic revelation (concealment) may occur only when the entrants prod-
uct is of higher (lower) quality than the incumbents product. This result suggests that
revealing or concealing business model innovations is an important strategic decision for
innovators. Moreover, the result implies that there may be a range of business model innovations that end up not being implemented because of the expected competitive imitation
by incumbents.
We also find that not all business model combinations emerge in equilibrium. In particular, we find that the equilibrium industry configuration often entails both firms competing
through different business models (and our model, therefore, provides a rationale for firm
heterogeneity). Understanding why firms operate under different business models in the
same industry is a long-standing question in strategy. Our study shows that explicitly endogenizing firms choices of business models in a game-theoretical framework is a promising
approach to tackle this question.
While much of the prior literature on positioning focuses on differentiation through product design, our work finds that the benefits from business model differentiation (i.e., the value
of innovation) can be substantial for both firms: indeed, we find that both the incumbent
and the entrant could make more profits with the innovation, even if the incumbent does
not directly benefit from it.
Finally, we find that the value of business model innovation may be so substantial that
the incumbent may prefer a duopoly than remaining a monopolist. This happens when the
entrants choice of business model reveals the innovation, and the benefit to the incumbent
from learning about it more than compensates for the loss of profit incurred in competing
with the entrant. To the best of our knowledge, our paper is the first to show formally that,
when competitors complement each other through business model innovation, competition
for the same customers with vertically differentiated products can lead to more profits for
the incumbent than a monopoly.
Related literature
Our study contributes to several strands of literature. First, our paper contributes to the literature on innovation and imitation. Scholars have looked at how new products or processes
can be imitated by their competitors (e.g., Benoit 1985; Gallini 1992; Pepall and Richards
1994; Ethiraj, Levinthal, and Roy 2008) and whether firms should disclose or license their
innovations (e.g., Hill 1992; Gans, Murray, and Stern 2008; Mukherjee and Stern 2009). Following these studies, we assume firms can imitate each others business model innovations
once they are revealed.
The literature on imitation is closely related to the broader literature on the transfer
of best practices among firms (e.g., Csaszar and Siggelkow 2010). Studies have identified a
variety of factors such as the absorptive capacity of the imitator, and the complexity of the
strategy to be imitated that facilitate or inhibit successful transfer of practices across firms
or across different units within a firm (e.g., Cohen and Levinthal 1990; Kogut and Zander
1992; Zander and Kogut 1995; Rivkin 2000; Knott 2003; Szulanski, Cappetta, and Jensen
2004). Unlike these studies that typically look from the perspective of an imitating firm and
examine practices within it, we study an innovators decision on whether to reveal its new
way of creating and capturing value, and explicitly model the competitive dynamics between
the innovator and its imitator in a game-theoretic framework. In addition, we also allow the
imitator to creatively combine the innovators new business model with its current model to
create new ones.
Our work also contributes to the growing literature examining competitive dynamics
between firms with different business models. Studies have examined such dynamics in a
number of industries, including the software industry (e.g., Casadesus-Masanell and Ghemawat 2006; Economides and Katsamakas 2006; Casadesus-Masanell and Yoffie 2007; Lee
and Mendelson 2008), the cable industry (Seamans forthcoming), and the music industry
(Casadesus-Masanell and Hervas-Drane 2010). Much of this literature has focused on interactions between firms with exogenously given business models. A few recent studies (e.g.,
Casadesus-Masanell and Zhu 2010; Casadesus-Masanell and Llanes 2011; Lin et al. 2008)
have endogenized firms choices of business models by allowing firms to select their business
models before deciding their optimal tactics to compete (as they would do in the real world).
In these studies, however, the available business models are assumed to be common knowledge among all firms. As a result, these models are not suitable for evaluating the value of
business model innovation. Our paper extends this line of work by explicitly modeling the
role of innovation. We assume that, initially, an entrant is the only firm aware of a new
business model and it chooses its business model before the incumbent reacts. The set of
business models available to the incumbent may expand or remain unchanged, depending on
whether the entrant chooses to reveal the innovation.
Our approach to modeling innovation follows the literature on unawareness (see Dekel,
Lipman, and Rustichini 1998, and references therein). Bages-Amat (2008) suggests that
incorporating unawareness is a useful approach to studying innovation and creativity, and
(according to this literature, e.g., Li 2008; Bages-Amat 2008), an agent is unaware of something if he does not know it, and he does not know that he does not know it (and so on ad
infinitum). Therefore, being unaware of a business model is different from simply not
knowing about it: a firm that does not know the new business model, but is aware of it (i.e.,
knows that it does not know about it) can still take it into account. We apply the concept
of unawareness to study business model innovation. In our game, the incumbent is bounded
rational: if the entrant does not reveal the business model innovation, the incumbent remains
unaware of the innovationand continues to compete as if the new business model did not
exist; once the entrant reveals the innovation, however, the incumbent becomes aware of it,
and can then take it into account when formulating its strategy. As Gavetti and Levinthal
(2004) point out, bounded rationality in theoretical models has typically taken the form of
myopic hill-climbing, or quasi-Skinnerian bases of action. Strategic action clearly involves
greater degrees of intentionality, so fuller representations of cognition would need to be incorporated into such theoretical efforts. Our approach to incorporating unawareness allows
10
the players to have great degrees of intentionalitythe incumbent always makes an optimal
decision given the information it has and becomes fully rational once the entrant reveals the
innovation.
Our approach to modeling competitive dynamics between the entrant and the incumbent
is similar to Shaked and Sutton (1982) in that firms products are vertically differentiated and
the extent of differentiation affects the intensity of competitive rivalry. Different from Shaked
and Sutton (1982), our setting also allows firms to differentiate themselves by adopting
different business models. The difference between our setup and that of the Shaked and
Sutton (1982) is most pronounced when we consider two firms offering products of similar
quality. In their setup, the low product differentiation leads to intense price competition
and nullify both firms profits. Hence, firms find it hard to co-exist. In our model, the two
firms can easily co-exist and be profitable by choosing different business models, which create
differentiation in the realized product quality.
Finally, the paper is related to the literature on platforms and multi-sided markets (e.g.,
Rochet and Tirole 2003; Caillaud and Jullien 2003; Armstrong 2006; Hagiu 2009; CasadesusMasanell and Ruiz-Aliseda 2009; Zhu and Iansiti forthcoming). Platforms are institutions
that act as intermediaries to enable transactions between multiple sides of a marketsuch
as sponsors and consumers. Most of the literature on multi-sided markets and platforms
considers situations where the sides attract one another. For example, operating system
platforms connect two sides: independent software vendors and users. Clearly, the more
applications are available for a particular operating system, the more attractive that system
is to users. Likewise, the larger the number of users of a particular operating system, the
more attractive it is for developers to produce applications for that system. When a platform
is sponsor-based, however, consumers prefer to access the product without interference by
the sponsors (i.e., consumers prefer books or email without ads, porn without captchas, or
access to slopes without sales pitches). Our paper contributes to this literature by exploring
the desirability (or lack thereof) of entering a market with a two-sided business model when
11
one side imposes a negative externality on the other.

The rest of the paper is organized as follows. We first present our model setup and
provide the theoretical results. We then discuss the implications of the results and conclude
after suggesting some extensions to the analysis.
MODEL
Setup
Our model involves two profit-maximizing firms, an incumbent and an entrant, indexed by
k {i, e}. Firms may obtain revenues from two different sources: they may charge a price for
their product pk , and they may find a business model innovation that allows them to monetize
their offering while giving the product away to customers. To model the second source of
revenue, let mk be the intensity of this alternative monetization effort (a variable that can be
decided by firms within the game, since such choices are endogenous within its framework).
In the examples above, mk is the number of ads in the book or on the website, the number
of captchas to be solved to access free porn, and the number of hours that skiers must listen
to promoters of timeshare condominiums. If sk (mk ) individuals adopt the product, then
the total revenue derived from this monetization effort mk is: Mk = sk (mk ) mk , where
> 0 is an exogenous parameter that represents the (per consumer) monetization rate.4,5
For simplicity, we assume that firms face no capacity constraints.
On the demand side, there is one unit mass of consumers. Consumers are differentiated
by their type , which represents their marginal willingness to pay for product quality and is
uniformly distributed on [0, 1]. The utility that a consumer of type receives from product
k {i, e} is U () = (qk m2k ) pk , where qk > 0 denotes the (exogenous) quality of product
4
In advertising models, > 0 corresponds to the (per consumer) advertising rate charged to each
advertiser. See, for example, Gabszewicz, Laussel, and Sonnac (2004).
5
We note that it would be natural to allow the composition of the firms customer base to affect sponsors
willingness to pay for access to the firms customers. The simplest way to incorporate this dependence
would be by allowing to vary across consumer type. Unfortunately, such an extension renders the model
intractable. Thus, we follow earlier theoretical work (e.g., Armstrong 2006; Gabszewicz et al. 2004; Jiang
2010; Lin et al. 2008) and assume that is constant.
12
k. Note that the monetization intensity mk is felt as a nuisance by consumersthe greater

it is, the lower they perceive the quality of the firms product to be. The convex functional
form implies that mild monetization efforts (such as forcing consumers to listen for a few
minutes the selling of a timeshare apartment opportunity) are tolerated well but more intense
monetization efforts are irritating. Note that the use of a sponsor-based business model has
implications not only concerning value capture (sponsors, rather than consumers, are the
sole source of revenue), but also relating to value creation (the firms monetization efforts
lead to lower willingness to pay for the product).
Prior to entry, the incumbent employs the traditional business model: it sells its product
at positive prices but is unaware of other opportunities to further monetize its offering.
Thus, the traditional business model has: mk = and pk > 0.6 The entrant may adopt
the traditional business model or it may innovate as already described. If it innovates, its
business model has: mk > 0 and pk = . Thus, the business model innovation is 180 away
from the traditional business model. We denote the traditional business model by T and the
new business model (the innovation) by I.
If entry occurs and the entrant innovates, the incumbent becomes aware of the entrants
monetization idea and may then choose to reconfigure its business model to respond. For
example, the incumbent may choose to compete with I, which entails moving from (mk =
, pk > 0) to (mk > 0, pk = ). Alternatively, it may decide to stay put with the original
model T , or to adopt elements of the new business model by setting (mk > 0, pk > 0) so that
its product is sold at positive prices and, at the same time, it is further monetized through
mk . We denote this combined business model by M , because it is a mixture of T and I.
We impose a non-negativity constraint on price pk and monetization intensity mk , and
normalize consumers utility from outside options to zero. Each consumer only adopts one
product. In addition (and without loss of generality), we adopt two tie-breaking rules: 1)
if a consumer receives zero utility from adopting a product, they will choose to adopt the
6
We use the empty set symbol to denote that that element is not in the business model.
13
product, and 2) if a consumer is indifferent between two products, they will choose to adopt
the product with a higher quality.
Timing
The timing of the game is as follows. First, the entrant decides its business model (the
incumbents model is preset at T ). Second, the incumbent learns about the entrants choice
of business model and chooses a business model to respond. Third, tactical choices (prices
and/or the monetization intensities) are made by both firms, and demand and profits are
realized.
As mentioned, the set of business models available to the incumbent in the second stage
is contingent on the entrants business model choice. If the entrant chooses T , then the
incumbent remains unaware of how to further monetize the producttheir choice is limited
to T . However, if the entrant chooses I, then the set of business models available to the
incumbent expands.
We allow the incumbent to respond to entry in one of two waysby staying with one
product and competing through business models T , I, or M , (I and M are available only
if the entrant has chosen I), or by introducing a fighting brand.7 We use a 2-tuple, (x, y),
to denote the business model choices of the two firms: x denotes the entrants and y the
incumbents. Different business model choices give rise to different tactical interactions
between firms, which we study in the next section.
Figure 1 shows the game where each combination of business models corresponds to a
different subgame. The possible business model combinations we consider are:
(T, T ): The entrant chooses the traditional business model and the incumbent responds with
the same business model;
(T, T T ): The entrant chooses the traditional business model and the incumbent responds
7
The term fighting brand has been frequently used by scholars and practitioners (e.g., Mintzberg
1987; Rao, Bergen, and Davis 2000) to refer to situations where an incumbent responds to competition by
expanding its product line with a lower-quality product.
14
by introducing a fighting brand with quality qi0 < qi . Therefore, the incumbent offers two
products in this case, the original product of quality qi and the fighting brand of quality qi0 ;8
(I, T ): The entrant chooses to innovate and the incumbent responds by staying put with the
traditional business model;
(I, M ): The entrant chooses to innovate and the incumbent responds with the mixed business
model (combining price and monetization efforts);
(I, I): The entrant chooses to innovate and the incumbent imitates and responds by imitating
this new business model;
(I, T T ): The entrant chooses to innovate and the incumbent responds by introducing a
fighting brand with quality qi0 < qi based on the traditional business model. The incumbent
winds up with two products;
(I, T I): The entrant chooses to innovate and the incumbent responds with a fighting brand
with quality qi0 < qi based on the new business model. The incumbent winds up with two
products.
[Figure 1 about here.]
It is easy to see that business model II (where the incumbent offers two products, both
based on the innovation) is dominated by I when the entrant chooses I. Moreover, business
models IT and T I (where the incumbent offers two products, one based on the traditional
business model and the other on the innovation) are equivalent. Therefore, it is unnecessary
to consider subgames (I, II) and (I, IT ) explicitly in our setup and this is why they do not
appear in Figure 1.
Business models as profit functions
As noted above, by business model we mean the logic of the firm, the way it operates and
how it creates and captures value for its stakeholders (Baden-Fuller et al. 2008; Casadesus8
We model the choice of qi0 as preceding the pricing choices of the two firms, since quality decisions are
often longer-term decisions than price decisions.
15
Masanell and Ricart 2010). To be able to work formally with business models, we represent
them in the form of simple profit functions. Thus, the choice of a particular business model
corresponds, in our development, to the choice of a particular profit function. Profit functions are highly simplified, reduced form representations of business models. These stylized
representations allow tight mathematical analyses. Zott and Amit (2010) propose the use
of Porters (1996) activity systems to represent business models. Porters activity systems
embody richer representations of business models and provide a textured picture of how the
firm creates and captures value. Activity systems emphasize that a firm is more than the
mere addition of activities as complementarities may result in important competitive advantages. On the negative side, activity systems are not amenable to game-theoretical analysis
because they often contain many elements and are too complex.9
To illustrate our approach, consider the subgame (I, T ) where the entrant has chosen
to innovate and the incumbent stays put with the traditional business model. The profit
functions are obtained as follows. We first examine the case where qi > qe . Here, the
incumbent maximizes profits by setting pi and the entrant maximizes profits by setting me
subject to the constraint that qe m2e 0 (so that its product has nonnegative net quality).
As the entrants product is given away for free, consumers who do not buy product i will
adopt product e. The type of the consumer who is indifferent between the two products, ,
is defined by qi pi = (qe m2e ) and the profits are:
(I,T )
i
9
= (1 ) pi
and e(I,T ) = me ,
(1)
We think of profit functions as representations of business models as if looked at from a distance. We

could zoom down closer to the actual details of the business model used by the firm and come up with
more complex profit functions that explicitly accounted for additional elements in the business models that
we have not considered. For example, the particular human resource management policies in place, the
production technologies used, or the marketing policies (just to name a few) are all part of the logic of the
firm, the way it operates and how it creates value for its stakeholders and, thus, are all part of a firms
business model, and could be included in the profit function to have a more detailed representation of the
firms business model. However, in most cases, these closer, more complete representations of business
models are too complex to be amenable to mathematical analyses.
16
subject to 0 1, and qe m2e 0.

Consider now the case where qe > qi . There are two cases:
Case 1: qe m2e qi . Here, as the entrant product is free and is of higher quality, the
incumbent is pushed out of the market, and all consumers adopt product e. The profits of
(I,T )
the two firms are: i
(I,T )
= 0 and e
= 1 me = me .
Case 2: qe m2e < qi . Now, as the entrants product is free, all consumers will adopt either
the incumbents product or the entrants product as long as qe m2e 0. Profit functions
in this case coincide with equation (1).
The entrant will compare the profits from these two cases and decide the optimal level of
me . This concludes our derivation of the profit functions for subgame (I, T ). The derivations
of profit functions for the six remaining subgames are presented in Appendix A of the working
paper version of this paper (Casadesus-Masanell and Zhu 2011).
RESULTS
We look for the subgame perfect equilibria of the game. Therefore, we proceed by backward
induction by first finding the equilibrium tactical choices in each subgame (pk and/or mk ,
depending on the business model combination) and then considering the equilibrium business
model choices.
Optimal tactics for each business model combination
The equilibrium analysis of the optimal tactical choices for each subgame is straightforward.10
We allow the entrants product quality, qe , to be higher or lower than the incumbents,
qi . In both cases, we find that if the entrant chooses the traditional business model, both
firms will coexist in equilibrium. As the incumbent does not learn about the innovation, it
will continue to adopt the traditional business model.
If the entrant chooses to innovate in its business model, it is possible that the entrant or
10
See Appendix B of the working paper version of this paper for the equilibrium analysis (CasadesusMasanell and Zhu 2011).
17
the incumbent will be pushed out of the market. In particular, when the entrants product
is of higher quality than the incumbents, the incumbent can co-exist with the entrant only
when its product is of relatively high quality (qi >
3
4
qe ). In this case, the entrant chooses
to maximize monetization intensity such that its net quality, qe m2e , is 0, as pushing the
incumbent out of the market requires setting a low monetization intensity which brings lower
profits. When the entrants product is of lower quality than the incumbents, the entrant
risks being pushed out of the market by the incumbent when is large.
These results suggest that an entrant that does not innovate cannot be killedone that
does, however, can be forced out if it introduces the innovation in the wrong situation.
Therefore, if the entrant is unsure about how its products quality compares to that of the
incumbent, it is safer to choose not to innovate but, instead, to compete through the
traditional business model.
Business model choice
Having derived the equilibrium tactics and payoffs under each possible combination of business models, we now analyze the first and second stages of the overall gamewhen firms
choose their business models.
The following diagram shows the equilibrium business model combinations for different
parameter values. The horizontal axis is the monetization rate and the vertical axis is the
quality ratio qe /qi .
Proposition 1 summarizes the features of Figure 2a and Proposition 2 those of Figure 2b.
Proposition 1. When qi < qe :
a. The only business model combinations that may arise in equilibrium are (T, T ), (I, T ),
and (I, T T ).
b. The entrant always survives and the incumbent is pushed out in the (I, T ) region only.
18
c. (T, T ) is the equilibrium business model configuration when is small and qe /qi is
large. (I, T T ) is the equilibrium business model configuration when is large and qe /qi
is small. (I, T ) is the equilibrium business model configuration when both and qe /qi
are large.
To understand why (I, I), (I, T I), (I, M ), and (T, T T ) cannot be equilibrium business
model configurations, note that when the entrant chooses to innovate and qe > qi , the
incumbent will never respond with a business model that involves mi > 0 because the
Bertrand-style competition results in qe m2e = qi , pushing the incumbent out of the market:
hence (I, I) and (I, T I) will never be the equilibrium business model choices. Choosing a
mixed model can also never be an incumbents best response to fight an entrant adopting
the new business model, since the entrant will always prefer to push the incumbent out of
the market in this case. Finally, when the entrant chooses the traditional business model,
the only possible choices for the incumbent are T and T T , and as argued above, T T will be
dominated by T .
Part (b) states that the incumbent is pushed out of the market only when the equilibrium
business model configuration is (I, T ). Obviously, when the equilibrium business models are
(T, T ) both firms co-exist. In the (I, T T ) case, the incumbent best-responds to the entrants
adoption of the innovation by introducing a fighting brand that induces the entrant to
respond with a large monetization intensity. As vertical differentiation between the entrant
product and the incumbents high-quality product increases, the incumbent can earn greater
profits: however, the introduction of a fighting brand in response to an entrant adopting
innovation results in positive incumbent profits only when the quality ratio qe /qi is less than
4/3.
To understand part (c), note that when is large, the entrant has a strong incentive
to innovate. In this case, if the quality ratio qe /qi is larger than 4/3 the entrant pushes
the incumbent out of the market no matter which business model the incumbent adopts.
However, when the quality ratio is lower than 4/3, the incumbent can avoid being killed by
19
choosing T T , which forces the entrant to set the maximum possible monetization intensity.
Vertical differentiation increases and the incumbent earns positive profits. When is low and
the quality ratio qe /qi sufficiently large, the entrant prefers to choose the traditional model
as the monetization revenue is low otherwise. In this case, the incumbent (still unaware of
the innovation) can only respond by choice of T or T T (as it continues to be unaware of the
innovation) and, as we have argued, the former dominates the latter.
We now turn to studying the situation where qe < qi , as illustrated in Figure 2b.
Proposition 2. When qi > qe :
a. The only business model combinations that may arise in equilibrium are (T, T ), (I, T ),
and (I, M ).
b. Both firms co-exist in equilibrium.
c. (I, M ) is the equilibrium business model configuration when is intermediate and qe /qi
is small, and (I, T ) is the equilibrium business model configuration when both and
qe /qi are intermediate. For all other values of and qe /qi , (T, T ) is the equilibrium
business model configuration.
The intuition for why (I, I) and (I, T I) are never equilibrium outcomes is the same as
in Proposition 1, except that here it is the entrant rather than the incumbent that tries to
avoid being killed. (I, T T ) and (T, T T ) do not occur in equilibrium because having a second,
low-quality product makes no difference to the incumbents profits.11
To understand part (b), note that the incumbent cannot be pushed out because it has a
higher-quality product and can always price its product close to zero to obtain positive demand. Moreover, the entrant can always choose to compete through the traditional business
model to obtain positive profits.
11
It is important to realize that the incumbent is not concerned about being pushed out of the market
when qe < qi . Therefore, releasing the second product does not serve the same purpose as when qe > qi (see
the discussion following Proposition 1).
20
Turning to (c), note that (I, M ) is the equilibrium business model configuration that has
a quality ratio below 1/2, in which case, the entrant is pushed to the corner where its net
quality is zero (i.e., me = qe ). When the entrant adopting the innovation is expected to
be at the corner, the incumbents optimal choice is the mixed business model.
In the (I, T ) region, the monetization rate is sufficiently high for the entrant to prefer
to innovate but not sufficiently high for the incumbent to best-respond by imitating. Nor
will the incumbent choose the mixed model as its response because when qe /qi > 1/2,
monetization intensities are strategic substitutes. This means that if the incumbent chooses
the mixed business model M (thus setting mi ), product differentiation will be smaller than
if it competes with the traditional model (because the entrant will respond to M by setting a
lower me ). The reduced product differentiation is more detrimental to the incumbents profits
than the additional monetization of the customer base (given the relatively low monetization
rate).
To understand the (T, T ) region, note first that when is large, the entrant does not want
to reveal the innovation because the incumbent will respond by imitating the new business
model and will end up pushing the entrant out of the market. If, however, the monetization
rate is low, the business model innovation is unattractive because the impact of additional
mk on market share is substantial but the additional monetization would be very small. In
this case, the entrant chooses the traditional business model and the incumbent responds
with the same business model.
DISCUSSION
This section discusses several implications of our formal business model innovation framework.
21
Strategic revelation and strategic concealment

We begin by examining the conditions under which the entrant prefers to strategically reveal
or conceal its innovation. We first formally define strategic revelation and concealment in
our context.
Strategic revelation refers to a situation where the entrant prefers to compete through
the new business model when it would choose not to do so if the incumbent was
expected to continue competing through the traditional model.
Strategic concealment refers to a situation where the entrant prefers to not compete
through the new business model when it would choose to do so if the incumbent was
expected to continue competing through the traditional model.
In other words, strategic revelation means choosing the new business model (revealing the
innovation) to induce the incumbent to change its business model in a way that is beneficial
to the entrant; and strategic concealment means choosing the traditional business model
(concealing the innovation) to prevent the incumbent from changing its business model in a
way that is detrimental to the entrant.
Figures 3 and 4 illustrate the regions where the entrant strategically reveals or conceals
its business model innovation, respectively. In both figures, panel (a) shows the entrants
optimal choice of its business model under the assumption that the incumbent will not
change its business model.12 Panel (b) superimposes panel (a) and the equilibrium business
model configurations derived in Figure 2 to show the regions with strategic revelation and
concealment.
Region 1 in Figure 3b exhibits strategic revelation: here the entrant reveals the innovation
by choosing to compete through the new business model to induce the incumbent to change
12
Of course, we are allowing the incumbent to adjust price in response to the entry.
22
its business model. Specifically, because region 1 is contained in region (I, T T ), the entrant
knows that the incumbent will respond by introducing a fighting brand, which in turn will
induce the entrant to increase its monetization intensity that help increase entrants profits.
Interestingly, if the incumbent continued to compete through the traditional model without
a fighting brand, in equilibrium, the entrant would lower monetization intensity and thus
its profits. The reason is that the trade-off between market share and monetization revenue
per user is resolved differently when the fighting brand is availablesince the fighting brand
makes the entrant less worried about market share and induces it to choose more aggressive
monetization me .
Turning to the case where qe < qi , regions 1 and 2 in Figure 4b exhibit strategic concealment: the entrant chooses the traditional business model, thus concealing the innovation,
even though it would choose to reveal it if the incumbent were constrained to stay with the
traditional model.
Strategic concealment takes place through different mechanisms in these two regions.
In region 1, if the entrant chose to innovate in its business model, the incumbent would
respond by introducing a fighting brand based on the new model. Because the entrant has
a product of lower quality, it would be pushed out of the market, so it chooses to conceal
the innovation. In region 2, if the entrant chose to innovate, the incumbent would respond
by adopting the mixed business model. The resultant lower vertical differentiation would
end up hurting (though not killing) the entrant, so here, again, the entrant is better off by
concealing the innovation.
Figures 3 and 4 indicate that strategic revelation (concealment) may occur only when the
entrants product is of higher (lower) quality than the incumbents product. This suggests
that business model innovations are more likely to be revealed when the quality of the
innovators product is high and that revealing or concealing business model innovations is
23
an important strategic decision for innovators. Moreover, the result implies there may be
a range of business model innovations that end up not being implemented because of the
expected competitive responses by incumbents.
Who benefits from business model innovation?
To tackle this question, we consider as a benchmark a situation where the entrant has not
come up with the business model innovation. This means that the entrant must compete with
the traditional business model and we compute the profits that the incumbent and the entrant
obtain if the entrant adopts the traditional business model. We compare this benchmark
to the profits that both firms obtain under the equilibrium business model configurations
that we have derived in the section titled Results (which are summarized in Figure 2).
This comparison allows us to determine the value of the business model innovation to the
incumbent and to the entrant.
Proposition 3. The entrant strictly benefits from the innovation in all regions where it
adopts the innovation (i.e., (I, M ), (I, T ), and (I, T T ) in Figure 2). The incumbent strictly
benefits from the innovation in the (I, T T ) region of Figure 2a, and (I, T ) and (I, M ) region
of Figure 2b, and becomes strictly worse off in the region (I, T ) of Figure 2a. The innovation
has no effect on either the incumbents or entrants profitability in regions where the entrant
does not adopt the innovation (i.e., the (T, T ) regions of Figure 2).
Obviously, if the entrant does not adopt the innovation, both the entrant and incumbent
will compete with the traditional business model and, as a result, their profitability will
remain unchanged. Because the entrant is the first mover and has superior knowledge from
the innovation, it will benefit from choosing the new business model (i.e., regions (I, M ),
(I, T ), and (I, T T ) in Figure 2)otherwise it would choose to compete with the traditional
business model.
In regions (I, T T ) of Figure 2a and (I, M ) of Figure 2b, the entrant will be at the corner,
so the incumbent makes more profits than what it would have if the entrant had adopted
24
the traditional business model (as in this case the entrant would not be at the corner). In
region (I, T ) of Figure 2b, with innovation, the entrants net quality becomes lower and the
two products are more differentiated. With the low intensity of competition, the incumbent
profits increase with the innovation. The incumbents profitability decreases, however, in
region (I, T ) of Figure 2a: without innovation, the two would co-exist with both earning
positive profitsbut with innovation, the entrant will push the incumbent out.
Might the incumbent ever prefer to be a duopolist rather than a monopolist?
We now investigate whether the benefits from innovation could ever be so substantial that the
incumbent preferred facing competition from an innovative entrant to remaining a monopolist. In this case, the benchmark is a situation where the incumbent faces no competition.
We show that while the transition from monopoly to duopoly generally implies lower profitability, in certain circumstances, an incumbent may strictly prefer a duopolistic industry
structure!
Proposition 4. The incumbent is strictly better off as a duopolist than as a monopolist in
region (I, M ) of Figure 2b, and is indifferent between being a duopolist and being a monopolist
in region (I, T T ) of Figure 2a. In all other regions in Figure 2, the incumbent is strictly
better off as a monopolist.
In all regions where the entrant chooses the traditional business model, the incumbent is
strictly worse off in a duopoly, as it does not learn the innovation and has to compete with a
product with positive net quality. Hence, we only need to consider the regions (I, M ), (I, T )
and (I, T T ) in Figure 2. In the (I, T ) region of Figure 2a, the incumbent is pushed out of the
market; in the (I, T ) region of Figure 2b, the incumbent is competing with an entrant whose
product has a positive net quality. Hence, the incumbent becomes worse off with the entrant
in these two regions. The incumbent makes the same profits with or without the entrant
in the region (I, T T ) of Figure 2a because in equilibrium, the net quality of the entrants
product is zero and thus it has no strategic interaction with the incumbents products.
25
Interestingly, the incumbent gains strictly greater profitability in the (I, M ) region of
Figure 2b. In this case, because the quality ratio qe /qi is low, the net quality of the entrants
product is zero. In addition, the incumbent now learns how to combine the innovation with
the traditional model and can derive some additional revenue (through the monetization of
its customer base). Therefore, in the (I, M ) region both firms benefit (strictly) from the
business model innovation!
We conclude that when firms compete with different business models, competition may
be beneficial to both firms, as well as to their consumers. This result contrasts with the
common scenario where firms compete with the same business model, when their profitability
is always adversely affected by competition. While this phenomenon has been documented
elsewhere through case study research (Casadesus-Masanell, Fernandez, and Jobke 2007;
Casadesus-Masanell and Campbell 2008), to the best of our knowledge, our paper offers the
first formal model where this sort of business model complementarity between competitors
arises endogenously.
LIMITATIONS, EXTENSIONS AND CONCLUSIONS

We conclude this paper by discussing several limitations and extensions to our model.
Endogenous quality
Our analysis has deemed the entrants quality qe exogenous.13 We now consider the endogenous choice of qe at a stage prior to the entrants choice of business model. Given that
the game is already intricate when quality is exogenous, we resort to numerical analysis to
analyze the entrants optimal quality level.
The game is as before except that there is a stage zero where the entrant chooses qe at
cost c(qe ) = c qe2 , where c > 0 is a constant cost parameter. To identify the entrants optimal
quality level, we plot the entrants maximum profits as a function of quality qe for different
13
We note, however, that the realized quality when the entrant chooses to innovate qe m2e is endogenous
as it depends on the monetization intensity chosen in equilibrium.
26
combinations of monetization rate and cost parameter c in Figure 5.14 The figure also
indicates the equilibrium business model choice by the entrant for each qe by making use of
the results summarized in Figure 2.15
The curve can be discontinuous as the incumbent may respond by changing its business
model. Comparing Figure 5a to Figure 5b, we find that when is lowso that there are
equilibria where the entrant chooses the innovative business model when qe < qi (see Figure
2b)the optimal quality is qe < qi and the business model chosen is the innovative one; when
is so high that the only equilibrium when qe < qi is (T, T ) (see Figure 2b), the entrant
prefers qe = qi and the innovative business model. Thus, while entrants with sponsor-based
business models will often choose to have lower-quality products than incumbents, they will
choose to increase their product quality when is high.
We also find that when c is low, as in Figure 5c, the entrants profits are maximal by
choosing quality qe > qi and competing with the traditional business model. When c is high
and low, however, the entrant will choose quality qe < qi and the traditional business
model (see Figure 5d).
These results show that there is no necessary correlation between equilibrium business
models and endogenous quality levels. The quality levels and business models observed
in equilibrium are the outcome of complex tradeoffs between the cost of quality c qe2 , the
monetization rate , and the expected competitive response by the incumbent as summarized
by Propositions 1 and 2.
Adoption cost of different business models
For simplicity, our model assumes the adoption cost of any business model to be zero, but
in reality, of course, there will always be cost involved in reconfiguring ones business model,
14
15
The figure has qi = 1.

The colorings of Figures 2 and 5 are mutually consistent.
27
which may vary for different business models. For example, ad-sponsored business models
could be less costly to adopt than fee-based models, since a firm such as Metro often does
not have to manage such complicated distribution channels, or maintain billing systems for
collecting subscription fees. Hence, our analysis may underestimate the benefits an entrant
might gain from the lower costs of its business model innovation. We also assume the costs
to the incumbent involved in learning about the business model innovation as zero, although
(as Rivkin 2000 shows) firms may suffer large penalties from small errors when learning and
imitating others business models. If the cost of imitation seems likely to be high and the
entrant product is of lower quality, the incumbent may choose not to learn and copy, and
the entrant will be more likely to reveal than to conceal its innovation.
Partial unawareness
In our analysis, we have assumed that the incumbent remains unaware of the business model
innovation if the entrant chooses not to adopt it. The assumption is consistent with prior
studies suggesting that business model imitation often requires changes to the entire activity
system and partial imitation may lead to large penalties (e.g., Zott and Amit 2010; Rivkin
2000). In this regard, our model and results are applicable to situations where an incumbent
knows about the new business model conceptually but is unable to adopt it without observing
the actual implementation by another firm.
When a business model innovation requires simple changes to an activity system, it is
possible that an incumbent can implement it by simply learning about the idea (through
employee mobility, for example) after an entrant has figured out the innovation (e.g., Agarwal, Ganco, and Ziedonis 2009). While the full analysis of the entrants strategic decisions
incorporating the possibility that an incumbent can implement a business model innovation
without observing its implementation is beyond the scope of the current paper, the spirit of
the paper carries over to this new setting: an entrant will face a similar strategic dilemma
over whether to engage in research on a new business model anticipating that an incumbent
may learn about it and implement it.
28
The possibility that an incumbent may adopt a business model innovation even if an
entrant does not adopt it does not necessarily lead to worse outcomes for the entrant. For
example, consider the situation where an entrants product quality is right below that of an
incumbent. In this case, our current analysis shows that {T, T } is the equilibrium business
model combination when is sufficiently large (see Figure 2b). In this case, the entrant
prefers to conceal the innovation as otherwise the incumbent would imitate and push the
entrant out of the market. The equilibrium profits for both firms under {T, T } are low due to
the low vertical product differentiation. In this case, both firms can earn higher profits if the
incumbent is able to learn about the innovation and implement it. In the new equilibrium,
the entrant adopts the traditional business model, the incumbent adopts the innovation, and
the two firms end up with products that are substantially vertically differentiated.
Empirical implications
While a few studies have examined the linkage between business model choices and firm
performance (e.g., Pauwels and Weiss 2008; Zott and Amit 2007, 2008), no empirical studies
have looked at how firms strategically conceal or reveal new business models. The lack
of empirical studies on this question is most likely due to the difficulty in collecting data
on possible business model implementations that did not happen. Our paper provides a
conceptual framework to think about this issue. While empirically testing our propositions
using a large sample dataset seems a daunting, if not impossible, task given the difficulty
involved in data collection, these theoretical results can be helpful to ethnographic researchers
as they conduct detailed case studies of business model innovations.
Our work also points to the possibility of a potential selection bias and an endogeneity
concern in empirical studies of business models. As firms may conceal new business models
because of competitive imitation, researchers will only be able to observe business models
adopted. In addition, the adoption decisions of business model innovations can be endogenously determined by firm-level characteristics and market factors (e.g., product quality and
the prevailing monetization rate in the market in our setting), both of which are likely to
29
be correlated with firm performance at the same time. Hence, empirical studies examining
business model choices and firm performance need to explicitly account for this endogeniety
to avoid spurious correlations.
Conclusion
Our paper shows that, as well as differentiating themselves in product quality terms, firms
can adopt different business models. We find some win-win scenarios, where entrants new
business models can benefit both them and their incumbent competitors, and is of greatest
value when both offer products of similar quality. Indeed, new entrants (such as Hotmail
and Pandora) offering similar products or services to the incumbents have became very
successful by adopting different business models.
Perhaps the most important implication of our study is that firms should take into account
the likely competitive effects before revealing a business model innovation. Entrants should
realize that incumbents will react to innovations in two main ways: they can keep their
business model intact and adjust its tactical variables (such as price); or they can adopt a
new business model so as to change their value creation and capture logic. The new business
model may be a replica of the innovators, or alternatively a new hybrid that combines some
of its elements with others from the incumbents original model in a mixed business model.
When an innovative entrant decides whether to adopt a new business model, it must
consider the possible responses of its rival. Such rival actions may nullify gains that might
otherwise accrue to the entrants innovation. The range and power of incumbents potential
strategic responses mean that there might be many business model innovations out there
that never see the light of day. Finding examples of strategic concealment is difficult because
we are looking for cases of business model innovations that have not been implemented by an
innovative entrant for fear of imitation. These are particularly difficult to find in the setting of
sponsor-based business models, which are a relatively new phenomenon. It is easy, however,
to observe situations where an entrepreneur mistakenly chooses to adopt a new business
model innovation and subsequently gets killed by an incumbent after the incumbent copies
30
the innovation. In such cases, the entrant would be better off concealing its business model
innovation. An example in the context of sponsor-based business models would be Metro
Spain, the first ad-sponsored free newspaper launched in Spain in 2001. After observing the
implementation of Metro Spain, several incumbents copied its business model. For example,
Recoletos launched Que! in 2005 and Editorial Pagina Cero launched ADN in 2006, both
of which are ad-sponsored free newspapers. In 2009, Metro Spain ceased its operation as a
result of stiff competition. The business model of Metro, the parent company of Metro Spain,
is also being imitated in many other countries (e.g., Switzerland) by incumbent newspapers.
As a result of this competitive imitation, Metros financial situation has been deteriorating
in recent years: its stock price dropped from more than 80 kr in 2000 to less than 0.8 kr in
2011.16,17
From the societys perspective, strategic concealment could generally be seen as not good
for welfare: some might argue that some form of intellectual property protection regulation
might encourage the emergence of more new business models. However, the practical difficulties of enshrining and policing such an extension to intellectual property rights would
probably be insurmountable.
Business model innovation is a slippery construct to study. The first implementation of a
new business model idea in an industry makes all firms in the sector (and beyond) aware of
the new way of conducting business, thus (there being no intellectual property protection)
limiting the innovators ability to take advantage of its idea. Hence, any studies of a business
model innovation that has already been implemented may offer little lessons for entrants.
While for readers of this paper the idea of sponsor-based business model innovation should
16
Source: http://www.reuters.com/finance/stocks/chart?symbol=MTROsdbb.ST, access August

2011.
17
While our study focuses on sponsor-based business models, entrepreneurs may find the concept of
strategic concealment useful when they consider adopting other business model innovations. For example,
PPG entered Chinas clothing retail industry in 2005 using a new business model that relies exclusively on
call centers and the Internet to sell traditional clothing. Traditional clothing retailers such as Younger and
Good News Bird in China responded by imitating the business model and integrating it with their existing
store-based business model, and pushed PPG out of the market in 2009.
31
be clear by now, the different ways in which such a general notion can come to life are
theoretically infinite, and even in practical terms myriad: every particular implementation
constitutes a business model innovation in its own right, to which our analysis can apply.
We can expect firms in all breeds of industries to continue to amaze us with unprecedented
new ways to capture value through sponsor-based business model innovation for many years
to come.
References
Adner, Ron, Rahul Kapoor. 2010. Value creation in innovation ecosystems: How the structure of
technological interdependence affects firm performance in new technology generations. Strategic Management Journal 31(3) 306333.
Agarwal, Rajshree, Martin Ganco, Rosemarie H. Ziedonis. 2009. Reputations for toughness in
patent enforcement: implications for knowledge spillovers via inventor mobility. Strategic
Management Journal 30(13) 13491374.
Amit, Raphael, Christoph Zott. 2001. Value creation in e-business. Strategic Management Journal
22(6-7) 493520.
Armstrong, Mark. 2006. Competition in two-sided markets. Rand Journal of Economics 37(3)
668691.
Baden-Fuller, Charles, Ian MacMillan, Benot Demil, Xavier Lecocq. 2008. Special issue call for
papers: Business models. Long Range Planning.
Bages-Amat, Arnau. 2008. Introducing creativity in game theory. Working paper.
Banbury, Catherine M., Will Mitchell. 1995. The effect of introducing important incremental
innovations on market share and business survival. Strategic Management Journal 16(S1)
161182.
Benoit, Jean-Pierre. 1985. Innovation and imitation in a duopoly. Review of Economic Studies
52(1) 99106.
Caillaud, Bernard, Bruno Jullien. 2003. Chicken & egg: Competition among intermediation service
providers. RAND Journal of Economics 34(2) 309328.
Casadesus-Masanell, Ramon, Neil Campbell. 2008. Betfair vs. the UK Bookmakers. Harvard
Business School Case and Teaching Note 709-417, 709-418.
Casadesus-Masanell, Ramon, Celso Fernandez, Moritz Jobke. 2007. Launching Telmore. Harvard
Business School Case and Teaching Note 708-414, 708-520.
32
Casadesus-Masanell, Ramon, Pankaj Ghemawat. 2006. Dynamic mixed duopoly: A model motivated by Linux vs. Windows. Management Science 52(7) 10721084.
Casadesus-Masanell, Ramon, Andres Hervas-Drane. 2010. Peer-to-peer file sharing and the market
for digital information goods. Journal of Economics & Management Strategy 19(2) 333373.
Casadesus-Masanell, Ramon, Gast
on Llanes. 2011. Mixed source. Management Science 57(7)
12121230.
Casadesus-Masanell, Ramon, Joan-Enric Ricart. 2010. From strategy to business model and onto
tactics. Long Range Planning 43(2-3) 195215.
Casadesus-Masanell, Ramon, Francisco Ruiz-Aliseda. 2009. Platform competition, compatibility,
and social efficiency. NET Institute Working Paper.
Casadesus-Masanell, Ramon, David Yoffie. 2007. Wintel: Cooperation and conflict. Management
Science 53(4) 584598.
Casadesus-Masanell, Ramon, Feng Zhu. 2010. Strategies to fight ad-sponsored rivals. Management
Science 56(7) 14841499.
Casadesus-Masanell, Ramon, Feng Zhu. 2011. Business model innovation and competitive imitation:
The case of sponsor-based business models. HBS Working Paper Number 11-003. Available
at http://www.hbs.edu/research/pdf/11-003.pdf.
Cohen, Wesley M., Daniel A. Levinthal. 1990. Absorptive capacity: A new perspective on learning
and innovation. Administrative Science Quarterly 35(1) 128152.
Csaszar, Felipe A., Nicolaj Siggelkow. 2010. How much to copy? Determinants of effective imitation
breadth. Organization Science 21(3) 661676.
Dekel, Eddie, Barton L. Lipman, Aldo Rustichini. 1998. Recent developments in modeling unforeseen contingencies. European Economic Review 42(3-5) 523542.
Economides, Nicholas, Evangelos Katsamakas. 2006. Two-sided competition of proprietary vs.
open source technology platforms and the implications for the software industry. Management
Science 52(7) 10571071.
Eisenhardt, Kathleen M., Behnam N. Tabrizi. 1995. Accelerating adaptive processes: Product
innovation in the global computer industry. Administrative Science Quarterly 40(1) 84110.
Ethiraj, Sendil K., Daniel Levinthal, Rishi R. Roy. 2008. The dual role of modularity: Innovation
and imitation. Management Science 54(5) 939955.
Gabszewicz, Jean J., Dider Laussel, Nathalie Sonnac. 2004. Programming and advertising competition in the broadcasting industry. Journal of Economics & Management Strategy 13(4)
657669.
Gallini, Nancy T. 1992. Patent policy and costly imitation. Rand Journal of Economics 23(1)
5263.
33
Gambardella, Alfonso, Anita M. McGahan. 2010. Business-model innovation: General purpose

technologies and their implications for industry structure. Long Range Planning 43(2-3) 262
271.
Gans, Joshua S., Fiona E. Murray, Scott Stern. 2008. Patents, papers & privacy: The disclosure
of scientific and commercial knowledge. Working paper.
Gavetti, Giovanni, Daniel A. Levinthal. 2004. The strategy field from the perspective of Management Science: Divergent strands and possible integration. Management Science 50(10)
13091318.
Hagiu, Andrei. 2009. Two-sided platforms: Pricing and social efficiency. Journal of Economics &
Management Strategy, 18(4).
Hill, Charles W. L. 1992. Strategies for exploiting technological innovations: When and when not
to license. Organization Science 3(3) 428441.
Jiang, Baojun. 2010. Ad-free versus ad-supported software platform. Working paper.
Katila, Riitta, Eric L. Chen. 2008. Effects of search timing on innovation: The value of not being
in sync with rivals. Administrative Science Quarterly 53(4) 593625.
Knott, Anne Marie. 2003. Persistent heterogeneity and sustainable innovation. Strategic Management Journal 24(8) 687705.
Kogut, Bruce, Udo Zander. 1992. Knowledge of the firm, combinative capabilities, and the replication of technology. Organization Science 3(3) 383397.
Lee, Deishin, Haim Mendelson. 2008. Divide and conquer: Competing with free technology under
network effects. Production and Operations Management 17(1) 1228.
Leiblein, Michael J., Tammy L. Madsen. 2009. Unbundling competitive heterogeneity: incentive
structures and capability influences on technological innovation. Strategic Management Journal 30(7) 711735.
Leiponen, Aija, Constance E. Helfat. 2010. Innovation objectives, knowledge sources, and the
benefits of breadth. Strategic Management Journal 31(2) 224236.
Li, Jing. 2008. Information structures with unawareness. PIER working paper 08-024.
Lin, Mei, Xuqing Ke, Andrew B. Whinston. 2008. Competitive strategies in business model innovation. Working paper.
Magretta, J. 2002. Why business models matter. Harvard Business Review (5) 8692.
Mintzberg, Henry. 1987. Five ps for strategy. California Management Review 30(1) 1124.
Mukherjee, Arijit, Scott Stern. 2009. Disclosure or secrecy: The dynamics of open science. Journal
of Industrial Organization 27(3) 449462.
Pauwels, Koen, Allen Weiss. 2008. Moving from free to fee: How online firms market to change
their business model successfully. Journal of Marketing 72(3) 1431.
34
Pepall, Lynne M., Daniel J. Richards. 1994. Innovation, imitation, and social welfare. Southern
Economic Journal 60(3) 673684.
Porter, Michael E. 1996. What is strategy? Harvard Business Review November-December
6178.
Rao, Akshay R., Mark E. Bergen, Scott Davis. 2000. How to fight a price war. Harvard Business
Review 78(2) 107116.
Rivkin, Jan W. 2000. Imitation of complex strategies. Management Science 46(6) 824844.
Roberts, Peter W. 1999. Product innovation, product-market competition and persistent profitability in the U.S. pharmaceutical industry. Strategic Management Journal 20(7) 655670.
Rochet, Jean-Charles, Jean Tirole. 2003. Platform competition in two-sided markets. Journal of
European Economic Association 1(4) 9901029.
Schroeder, Dean M. 2006. A dynamic perspective on the impact of process innovation upon competitive strategies. Strategic Management Journal 11(1) 2541.
Schumpeter, Joseph A. 1934. The Theory of Economic Development: An Inquiry into Profits,
Capital, Credit, Interest and the Business Cycle. Oxford University Press, London.
Seamans, Robert C. forthcoming. Fighting city hall: Entry deterrence and technology upgrades in
cable TV markets. Management Science .
Shaked, Avner, John Sutton. 1982. Relaxing price competition through product differentiation.
Review of Economic Studies 49(1) 313.
Shan, Weijan, Gordon Walker, Bruce Kogut. 1994. Interfirm cooperation and startup innovation
in the biotechnology industry. Strategic Management Journal 15(5) 387394.
Szulanski, Gabriel, Rossella Cappetta, Robert J. Jensen. 2004. When and how trustworthiness
matters: Knowledge transfer and the moderating effect of causal ambiguity. Organization
Science 15(5) 600613.
Teece, David. 2010. Business model, business strategy, and innovation. Long Range Planning
43(2-3) 172194.
Tirole, Jean. 1994. The Theory of Industrial Organization. MIT Press.
Zander, Udo, Bruce Kogut. 1995. Knowledge and the speed of the transfer and imitation of
organizational capabilities: An empirical test. Organization Science 6(1) 7692.
Zhou, Kevin Zheng, Fang Wu. 2010. Technological capability, strategic flexibility, and product
innovation. Strategic Management Journal 31(5) 547561.
Zhu, Feng, Marco Iansiti. forthcoming. Entry into platform-based markets. Strategic Management
Journal .
Zott, Christoph, Raphael Amit. 2007. Business model design and the performance of entrepreneurial
firms. Organization Science 18(2) 181199.
35
Zott, Christoph, Raphael Amit. 2008. The fit between product market strategy and business model:
Implications for firm performance. Strategic Management Journal 29(1) 126.
Zott, Christoph, Raphael Amit. 2010. Designing your future business model: An activity system
perspective. Long Range Planning 43(2-3) 216226.
36
Figure 1: The strategic innovation game.
#$$*$#%"
#$$($#%"
2.0
,$
1.0
*$
1.8
0.8
1.6
!&#$"%$
!"#$"%$
!&#$"%$
0.6
!"#$"%$
*(+$
0.4
1.4
)*+$
1.2
1.0
'$
0.0
($
0.2
!&#$""%$
0.1
0.2
0.3
(a) qe > qi .
!&#$'%$
0.4
0.5
!"
'*,$
0.0
)$
0.0
)$
0.1
0.2
0.3
(b) qe < qi .
Figure 2: Optimal business models.
0.4
0.5
!"
*(+$
37
#$$)$#%"
#$"%"#%"
2.0
*$
2.0
'"
1.8
1.8
!"#$"%$
1.6
1.6
1.4
1.4
$%&"
!&#$"%$
1.2
1.2
&"
1.0
'$
0.0
($
0.1
0.2
0.3
0.4
0.5
!"
')*$
(a) Optimal BM without imitation.
1.0
!"
0.0
#"
0.1
0.2
0.3
0.4
0.5
!"
!%'"
(b) Strategic revelation and concealment.
Figure 3: Optimal business models when qe > qi .
#$$'$#%"
#$!"!#%"
1.0
$!
1.0
)$
0.8
!%#$%&$
0.8
&"
0.6
0.6
)'*$
!"#$%&$
$"%!
0.4
0.4
0.2
0.2
0.0
($
0.0
($
0.1
0.2
0.3
0.4
0.5
!"
)'*$
(a) Optimal BM without imitation.
0.0
#!
0.0
#!
'"
0.1
0.2
0.3
0.4
0.5
!"
$"%!
(b) Strategic revelation and concealment.
Figure 4: Optimal business models when qe < qi .
(a) c = .10, = .20
(b) c = .10, = .35
(c) c = .06, = .20
(d) c = .085, = .01
.09
.06
.20
.20
.10
.085
.06
.35
.20
.01
Figure 5: Entrants profits as a function of qe .
ONLINE APPENDICES
APPENDIX A: PROFIT FUNCTIONS FOR THE SEVEN SUBGAMES
Subgame 1: (T, T )
In this case, the entrant chooses the traditional business model and the incumbent chooses
to compete with the same model.
We first consider the case when the incumbent has a better quality, i.e., qi > qe . The consumer 1 indifferent between purchasing from the incumbent and the entrant is determined
e
by 1 qi pi = 1 qe pe . Hence, 1 = pqii p
. The consumer 2 indifferent between purchasing
qe
from the entrant product and not purchasing at all is determined by: 2 qe pe = 0. Hence,
ii
2 =
pe
.
qe
Therefore, the profits for the incumbent and the entrant are:
(T,T )
e(T,T )
pi pe
) pi .
qi q e
pi pe pe
) pe .
=(
qi qe
qe
= (1
We can similarly derive the profit expressions when qe > qi :

(T,T )
e(T,T )
pe p i pe
) pi .
qe q i
qe
pe pi
= (1
) pe .
qe qi
=(
Subgame 2: (T, T T )
In this case, the entrant chooses the traditional business model and the incumbent offers two
products, i and i0 , with qi > qi0 . While the quality of qi is exogenously given, the incumbent
may create a fighting brand by strategically downgrading the quality of its original product,
choosing qi0 before engaging in the pricing subgame with the entrant.
When qi > qe , there are two cases to consider:
Case 1: qi > qi0 > qe . We first derive the types of indifferent consumers, 1 , 2 , and 3 ,
between adopting qi and qi0 , between adopting qi0 and qe , and between adopting qe and not
p p
p p
adopting. We have 1 = qii q i00 , 2 = qi00 qee , and 3 = pqee . Profits of the incumbent and the
i
i
entrant are:
(T,T T )
e(T,T T )
pi p i0
pi0 pe
pi pi0
) pi + (
) pi0 .
qi qi0
qi q i0
qi0 qe
pi0 p e pe
=(
) pe .
qi0 q e
qe
= (1
Case 2: qi > qe > qi0 . We again derive the types of indifferent consumers, 1 , 2 , and 3 ,
between adopting qi and qe , between adopting qe and qi0 , and between adopting qi0 and not
p p
p
pe
adopting. We have 1 = pqii q
, 2 = qee q i00 , and 3 = qi00 . Profits of the incumbent and the
e
i
i
entrant are:
(T,T T )
e(T,T T )
pi pe
pe pi 0
pi0
) pi + (
) pi0 .
qi qe
qe qi 0
q i0
pi pe pe pi0
=(
) pe .
q i qe
qe qi0
= (1
iii
When qe > qi , we must have qe > qi > qi0 . We could similarly derive the following profit
functions:
(T,T T )
e(T,T T )
pi pi0
pi0
pe pi p i pi0
) pi + (
) pi0 .
qe qi
qi qi0
qi qi0
qi 0
p e pi
) pe .
= (1
q e qi
=(
Subgame 3: (I, T )
Please see the main text as this case has been presented when we introduce the model in
detail.
Subgame 4: (I, M )
We first consider the case where qi > qe . The incumbent product now comes with monetization intensity, mi , and is priced at pi > 0. The indifferent consumer is defined by
pi
(qi m2i ) pi = (qe m2e ). Hence, = qi m2 q
2 , and the profits are:
e +m
i
(I,M )
pi
)(pi
2
qi mi qe + m2e
pi
me .
2
qi mi qe + m2e
qi m2i 0, qe m2e 0.
= (1
e(I,M ) =
s.t.
+ mi ).
For this business model to be meaningful, we need that pi > 0 and mi > 0. Otherwise, one
of the pure business models is the effective one.
We now consider the case where qe > qi . There are two cases.
Case 1: qe m2e qi . In this case, the incumbent is pushed out. The profits are thus:
(I,M )
= 0.
e(I,M ) = me .
s.t.
qe m2e qi .
Case 2: qe m2e < qi . In this case, the incumbent product will always be active as it can
come with lower monetization intensity and low price. The indifferent consumers type, ,
pi
is determined by (qi m2i ) pi = (qe m2e ). Hence, = qi m2 q
2 . The profits are
e +m
i
iv
thus:
(I,M )
pi
)(pi
2
qi mi qe + m2e
pi
me .
2
qi mi qe + m2e
qi m2i 0, qe m2e 0.
= (1
e(I,M ) =
s.t.
+ mi ).
The entrant will compare the profits from both cases and decide the level of me .
Subgame 5: (I, I)
When both the incumbent and the entrant provide free products, all consumers will adopt
the product with the highest net quality. This competitive situation is similar to Bertrand
competition, except that now the two firms are setting the monetization intensities, not
prices.
When qi > qe , the profits are:
(I,I)
e(I,I)
0
=
m
if qi m2i qe m2e
otherwise.
if qi m2i qe m2e
otherwise.
qi m2i 0 and qe m2e 0.
s.t.
When qe > qi , the profits are:
(I,I)
i
e(I,I)
s.t.
if qe m2e qi m2i
m otherwise.
i
m if q m2 q m2
e
e
i
e
i
=
0
otherwise.
Note that we are assuming that when both products, i and e, are of the same net quality,
consumers prefer the offering of the higher quality firm. This is without loss of generality.18
18
For example, when qi > qe , the incumbent could always set its net quality, qi m2i , at qe m2e + to
attract all consumers.
Subgame 6: (I, T T )
First, we consider the case where qi > qe . There are two cases.
Case 1: qi > qi0 > qe m2e . As the entrant product is free, as long as qe m2e 0,
all consumers will make adoptions. We first derive the types of the indifferent consumers
between the incumbents two products, and between the incumbent low-quality product and
p p
p
the entrant product: 1 = qii q i00 and 2 = q 0 qei0+m2 , respectively. The profits are thus:
i
pi p i0
pi0
pi pi0
) pi + (
) p i0 .
qi qi0
qi q i0
qi0 qe + m2e
pi0
e(I,T T ) =
me .
qi0 qe + m2e
s.t. qi > qi0 > qe m2e 0.
(I,T T )
= (1
Case 2: qi > qe m2e qi0 . In this case, product i0 obtains no demand in equilibrium. The
type of the indifferent consumer between product i and qe is = qi qpei+m2 . The profits are:
e
pi
) pi .
qi qe + m2e
pi
e(I,T T ) =
me .
qi qe + m2e
s.t. qi > qe m2e qi0 0.
(I,T T )
= (1
We now consider the case where qe > qi . There are three cases.
Case 1: qe m2e qi > qi0 . In this case, the incumbent is pushed out of the market and the
entrant has a demand of 1. The profits are thus:
(I,T T )
= 0.
e(I,T T ) = me .
s.t.
qe m2e qi .
Case 2: qi > qe m2e qi0 . In this case, the low quality product of the incumbent is pushed
out of the market. Hence, this is equivalent to the case where the incumbent uses only one
vi
product to respond to the entrant. From our discussion of the (I, T ) case, we have:
pi
) pi .
qi qe + m2e
pi
e(I,T T ) =
me .
qi qe + m2e
s.t. qi > qe m2e qi0 .
(I,T T )
= (1
Case 3: qi > qi0 > qe m2e . In this case, all three products may be active. We first derive
the types of indifferent consumers between the two products offered by the incumbent, and
p p
between the low-quality product of the incumbent and the entrant product as 1 = qii q i00
i
p
and 2 = q 0 qei0+m2 . The profits are thus:
i
pi pi0
pi p i0
pi0
) pi + (
) p i0 .
qi qi0
qi q i0
qi0 qe + m2e
pi0
me .
e(I,T T ) =
qi0 qe + m2e
s.t. qi > qi0 > qe m2e 0.
(I,T T )
= (1
In equilibrium, both the incumbent and the entrant will compare their profits from these
three cases and decide their optimal levels of qi0 , pi , pi0 , and me .
Subgame 7: (I, T I)
In this case, the incumbent introduces two products: product i that is traditional and product
i0 that is based on the new business model. When qi > qe , suppose that the monetization
intensities mi0 and me are such that the entrant is pushed out of the market. Then, consumers
either buy the high-quality product of the incumbent or consume the free product of the
incumbent. In this case, the indifferent consumer is determined by qi pi = (qi m2i0 ).
That is, = mpi2 . Suppose instead that the monetization intensities, mi0 and me , are such
i0
that the entrant is not pushed out of the market. Then, consumers either buy the highquality product of the incumbent or consume the product of the entrant. In this case, the
indifferent consumer is determined by qi pi = (qe m2e ). That is, = qi qpei+m2 .
e
vii
The profits are:

(I,T I)
e(I,T I)
s.t.
(1
pi
) pi
m2i0
(1 ) p
i
0
=
p i
m
pi
m2i0
e
qi qe +m2e
2
qi mi0 0, qe
mi0
if qi m2i0 qe m2e
otherwise.
if qi m2i0 qe m2e
otherwise.
m2e 0.
This business model is meaningful only when pi > 0, mi0 > 0, and me > 0.
Now consider the case where qe > qi . In this case, the entrant again engages in Bertrand
type competition with the free product of the incumbent. The entrant will push out both
products of the incumbent when qe m2e qi . The entrant will push out the free product
of the incumbent if qi > qe m2e > qi m2i0 . The case is equivalent to the (I, T ) case. When
qi m2i0 > qe m2e , the entrant product will be pushed out. In this case, the indifferent
consumer between the incumbents two products is determined by qi pi = (qi m2i0 ).
That is, = mpi2 . The profits are:
i0
(I,T I)
e(I,T I)
s.t.
0
if qe m2e qi
= (1 qi qpei+m2 ) pi
else if qi > qe m2e > qi m2i0
e
(1 pi2 ) pi + pi2 mi0 otherwise.

mi0
mi0
m
if qe m2e qi
e
= q qpi+m2 me else if qi > qe m2e > qi m2i0
e
i
e
0
otherwise.
APPENDIX B: OPTIMAL TACTICS FOR EACH BUSINESS MODEL COMBINATION

Propositions A-1 and A-2 describe the equilibrium tactics for each subgame.19 Without loss
of generality, we adopt the following tie-breaking rule: when two different business model
combinations yield the same payoffs, we select the combination with fewer products. This
rule captures the fact that introducing new products will generally involve some cost. For
example, Propositions A-1 and A-2 show that the payoffs are the same in subgames (T, T )
19
We provide all proofs in Appendix C.
viii
and (T, T T ). The tie-breaking rule suggests that (T, T ) is preferred to (T, T T ).
We consider first the interactions when the entrants quality is above that of the incumbent.
Proposition A-1. When qi < qe , the optimal prices and monetization intensities under
each business model combination are:
(T, T ): pi =
(qe qi )qi
4qe qi
and pe =
2qe (qe qi )
.
4qe qi
i )qi
(T, T T ): The optimal pi and pe are the same as in the (T, T ) case: pi = (q4qe q
and
e qi
(qe qi )qi0
2qe (qe qi )
pe = 4qe qi . The incumbent sets pi0 = 4qe qi in equilibrium such that there is zero
demand for product i0 .
(I, T ): pi = 0 and me =
qe qi . The incumbent is pushed out of the market.
(I, M ): pi = 0, mi = 0, and me =
market.
(I, I): mi = 0 and me =
qe qi . The incumbent is pushed out of the
qe qi . The incumbent is pushed out of the market.
(I, T T ): When qi < 34 qe , the entrant sets me = qe qi and the incumbent is pushed
out of the market. Otherwise, me = qe and the incumbent receives positive demand.
The equilibrium choices are qi0 = 0, pi = qi /2, and pi0 = 0. The incumbents lowerquality product i0 receives no demand but affects the equilibrium outcome.
(I, T I): pi = 0, mi0 = 0 and me =
market.
qe qi . The incumbent is pushed out of the
The intuitions for these results are as follows.

Subgame (T, T ). In this case, we obtain the standard result of two vertically differentiated
competitors fighting for the same customers.20 The high-quality firm sells twice the demand
(at a higher price) than the low-quality firm. In equilibrium, the entrant sells to more than
half of the market.
Subgame (T, T T ). Just as in the standard model of a muti-product monopolist with
vertically differentiated products, the incumbent prefers offering only one product: its lowquality product i0 is priced such that it obtains no demand, so payoffs in this case coincide
with those of (T, T ). Of course, if offering two products was costlier than offering only one,
(T, T T ) would be strictly dominated by (T, T ).
20
See, for example, Shaked and Sutton (1982) or Tirole (1994).
ix
Subgame (I, T ). In this case, the incumbent is pushed out of the market. The entrants
two choices are either to set the monetization intensity so that its products net quality is
greater than the incumbents quality, qe m2e qi , or it can choose a monetization intensity
so that its net quality is lower, qe m2e < qi . In the first case, the entrants product is both
free and of higher quality, so the incumbent is pushed out of the market. In the second
case, the entrant trades off market share and revenue per unit of share. The entrants profits
(I,T )
= qi qpei+m2 me .
are: market share ( qi qpei+m2 ) times revenue per consumer ( me ), i.e., e
e
e
As me increases, market share decreases faster than revenue per consumer grows. As a
consequence, the entrant will prefer to reduce me when qe m2e < qi . The entrant ends up
choosing me such that qe m2e = qi , and the incumbent ends up with no demand.
Subgame (I, M ). In equilibrium the incumbent is pushed out of the market for the same
reasons as in subgame (I, T ). Even though it has the option of using its monetization efforts
to differentiate itself from the entrant, the entrants trade-off between market share and
revenue per consumer still results in the incumbent being pushed out.
Subgame (I, I). In this case, there is Bertrand-type competition between the incumbent
and the entrant.21 Therefore, the entrant cannot choose a high me as it has to offer a product
of net quality at least as large as that of the incumbent without ads. Hence, the optimal
me is constrained by qe m2e qi . When this constraint is satisfied, all consumers adopt
product e. Therefore, it is in the entrants interest to maximize its monetization intensity
by setting me so that qe m2e = qi .
Subgame (I, T T ). In this case, there are three possible scenarios. First, the entrants net
quality may be larger than the incumbents high-quality product (i.e., qe m2e > qi ), and
the incumbent is pushed out because the entrants product is of higher quality and is free.
Second, the entrants net quality may be between the quality levels of the incumbents two
products (i.e., qi > qe m2e > qi0 ), making product i0 irrelevant. Just as in the (I, T ) subgame,
the entrants net quality ends up being equal to qi so the incumbent is again pushed out.
Finally, the entrants net quality may be lower than the incumbents low-quality product,
qe m2e < qi0 . Now, the same argument as in the (I, T ) subgame applies to the interaction
between the entrants product e and the incumbents low-quality product i0 . As in that case,
the entrant ends up setting its net quality to match qi0 , although, in this case, the incumbents
high-quality product obtains positive demand. It is optimal for the incumbent to set qi0 = 0
so that there is no interaction between its high-quality product and the entrants product,
leaving the incumbent with monopoly power for its high-quality product.
21
While in Bertrands model firms choose prices, here firms choose the monetization intensity.
The entrant thus has the power to decide whether the incumbent is pushed out or not (by
choosing me so that qe m2e > qi > qi0 , qi > qe m2e > qi0 , or qi > qi0 > qe m2e ).
The entrant can compare its profits in these scenarios: when the quality difference between
e and i is low (qi > 43 qe ), coexistence brings more profits to the entrant because (given
the incumbents relatively high quality) pushing the incumbent out requires setting a low
monetization intensity.
Subgame (I, T I). As in the (I, I) case, a Bertrand competition argument implies that
only the higher-quality product offered through I survives. Therefore, in equilibrium, the
incumbent will reduce the monetization intensity for its I product to zero. In the end
qe m2e qi = qi0 and hence, the incumbent obtains no demand.
We now present the results when the entrants product is of lower quality.
Proposition A-2. When qi > qe , the optimal prices and monetization intensities under
each business model combination are:
(T, T ): pi =
2(qi qe )qi
4qi qe
and pe =
qe (qi qe )
.
4qi qe
e )qi
(T, T T ): The optimal pi and pe are the same as in the (T, T ) case: pi = 2(q4qi q
and
i qe
qe (qi qe )
0
pe = 4qi qe . The incumbent sets qi0 = qi so that product i obtains zero demand.
(I, T ): When qi 2qe , pi = qi /2 and me =
me = qi qe .
qe ; when qi < 2qe , pi = qi qe and
(I, M ): We may have a corner solution in which qe m2e = 0 or an interior solution

in which qe m2e > 0.
At the interior solution, mi and me solve the following system:
q q + m2 = m3i
i
e
e
mi
p
m = q q m2
e
and pi = 12 ((qi m2i ) (qe m2e ) mi )) .

At the corner solution, mi solves: m3i + qi ( mi ) = 0, me =
mi ( + mi ))/2.
(I, I): mi =
qe , and pi = (qi
qi qe and me = 0. The entrant is pushed out of the market.
(I, T T ): Same as (I, T ).
xi
(I, T I): pi = 12 (qi qe + qi qe ), mi0 = qi qe , and me = 0. The entrant is

pushed out.
The intuitions for these results are as follows.
Subgames (T, T ), (T, T T ), (I, I), and (I, T I). The intuitions for these are analogous
to those in Proposition A-1, except that now the high-quality product is offered by the
incumbent.
Subgame (I, T ). The optimal tactics of the incumbent in this case depend on whether
the entrant sets its monetization intensity at the corner or not (i.e., me = qe or me > qe ),
which in turn depends on the exogenous vertical differentiation between the incumbents
and the entrants products. The entrants profits increase with its market share and with
me , but there is a tradeoff between the two. When the entrants product is of very low
quality (qi 2qe ), it is best for it to maximize monetization intensity because its market
i
share, = qi qep+m
2 , is insensitive to me (the derivative of with respect to me approaches
e
zero as the difference between qi and qe grows). On the other hand, if its quality is close to
the high-quality incumbent (qi < 2qe ), is sensitive to the monetization intensity and it is
optimal for the entrant to reduce me to gain market share.
When qi 2qe , there is no cannibalization between the two products: qe m2e = 0. The
indifferent consumer obtains zero utility. When qi < 2qe , the net quality of the entrant in
equilibrium is positive: qe m2e = qe qi /2 > 0. The indifferent consumer has positive utility
from both products.22
Subgame (I, M ). As the incumbent product is not free, consumers with low will not
buy it: they will adopt the entrants product as long as it offers positive utility. As a result,
both the incumbent and the entrant co-exist in equilibrium.
The solution may be at a corner, where the entrant chooses the maximum possible moneti
zation intensity (me = qe ) such that the utility for its product is zero. The corner solution
happens when the quality difference is large (i.e., qi > 2qe ). In this case, the unconstrained
p
profit-maximizing me (i.e., qi qe m2i ) would exceed the maximum monetization inten
sity that the entrant can possible have (i.e., qe ). The entrant chooses to set me at qe , and
the indifferent consumer receives zero utility. Or the solution may be interior (me < qe ).
It is interesting to note that = 1/2 in both cases. That is, the incumbent and the entrant always split
the market equally, regardless of their quality difference. Given any me , the residual demand for product i
is Di = 1 = 1 qi qpei+m2 . The marginal revenue implied by this demand function equals marginal cost
e
(which is zero) at Di = 21 regardless of the value of me . Of course, the equilibrium pi changes with me , and
so does the incumbent profits, but the equilibrium Di does not change.
22
xii
This happens when qi 2qe . In this case, the entrants product offers strictly positive utility.
The indifferent consumer thus gets positive utility.
Subgame (I, T T ). The entrant may be at the corner (me = qe ) or not (me > qe ). If
the entrant is at the corner, the incumbents and entrants products do not interact, so there
is no business stealing. In this case, the incumbent sets qi0 = qi and the outcome is the same
as in the (I, T ) subgame.
If the entrant is not at the corner, then the incumbent has two options as to the quality of
its fighting product. It can either set qi > qi0 > qe or qe > qi0 . The first case is never optimal
because the competition in the low end ends up hurting profits for the high-quality product,
and so the incumbent will seek to maximize profits by setting qi0 = qi . In the second case,
there are two possibilities. First, the incumbent introduces a very low quality i0 which does
not affect the entrants optimal amount of monetization intensity (as qe m2e > qi0 , and
i0 obtains no demand). This case is equivalent to subgame (I, T ). Second, the incumbent
introduces product i0 with quality close to qe . In this case, the entrant will set a smaller
monetization intensity to kill product i0 and the competitive pressure on product i will be
greater than if i0 had not been introduced.
In summary, the outcome of (I, T T ) coincides with that of the (I, T ) subgame.
APPENDIX C: PROOFS
Proof of Proposition A-1. Subgame (T, T ): The derivation is straightforward by setting the
i )qi
and
FOC of the profit function to zero and solve for the optimal pi and pe . pi = (q4qe q
e qi
pe =
2qe (qe qi )
.
4qe qi
(T,T )
The profits are i
qi qe (qe qi )
(4qe qi )2
(T,T )
and e
4qi2 (qi qe )
.
(4qi qe )2
Subgame (T, T T ): We first take FOCs of the profit functions and solve for the optimal pi ,
(qe qi )qi0
e (qe qi )
i )qi
pi0 and pe , assuming that qi0 is given. pi = (q4qe q
, pi0 = 4q
and pe = 2q4q
. The
e qi
e qi
e qi
(T,T T )
qi qe (qe qi )
profits are i
= (4qe qi )2 . Note that this profit is the same as in the (T, T ) case and is
(T,T )
independent of qi0 . Similarly, e

(T, T ).
4qi2 (qi qe )
.
(4qi qe )2
Therefore, this case is weakly dominated by
(I,T )
Subgame (I, T ): When qe m2e qi , the incumbent is pushed out the market and e
=
me . Hence the entrant will increase its monetization intensity as much as possible subject
(I,T )
(I,T )
to the constraint that qe m2e qi . Hence, me = qe qi . Thus, i
= 0 and e
=
(I,T
)
qe qi . When qe m2e < qi , the FOC of i
w.r.t. pi gives pi = 12 (qi qe + m2e ). The
(I,T )
FOC of e
w.r.t. me is negative. Hence, the entrant will choose me as small as possible
xiii
subject to the constraint, qe m2e < qi . At the end of the process, the entrant essentially
kills the incumbent. Therefore, the entrant will set me to qe qi and the incumbent is
pushed out.
(I,T )
(I,T )
In summary, the profits are: i
= 0 and e
= qe q i .
Subgame (I, M ): When qe m2e qi , the entrant pushes out the incumbent. Hence, the
(I,M )
(I,M )
optimal me = qe ql . We have i
= 0 and e
= qe q i .
(I,M )
When qe m2e < qi , the FOC of e
w.r.t. me is negative. Hence, similar to the (I, T )
case, the entrant will try to reduce me as much as possible. The incumbent responds by
lowering mi and pi . In the end, me = qe qi . The incumbent is again pushed out of the
market.
(I,M )
(I,M )
= 0 and e
= q e qi .
In summary, i
Subgame (I, I): In this case, the Bertrand-style competition happens and the entrant pushes
(I,I)
out the incumbent by setting qe m2e = qi , and obtain the whole market. Hence, i
=0
(I,I)
and e = qe qi .
Subgame (I, T T ): When qe m2e qi > qi0 , the incumbent is pushed out and the entrant has
(I,T T )
a demand of 1. The maximum me the entrant can set is qe qi . The profits are i
=0
(I,T T )
and e
= qe q i .
When qi > qe m2e qi0 , the incumbents product i0 is pushed out. me is bounded by
(I,T T )
(I,T T )
w.r.t. pi gives pi = 12 (qi qe + m2e ). The FOC of e
w.r.t.
qe qi0 . The FOC of i
me is negative. Similar to the (I, T ) case, product i will be pushed out as well. Hence, in
this case, the profits are the same as in the first case.
(I,T T )
When qi > qi0 > qe m2e , the FOC of e
w.r.t. me is negative so that the low
0
quality product of the incumbent, i , is pushed out and me = qe qi0 . It is optimal for
the incumbent to set qi0 = 0 and pi0 = 0 as in this case, me will be set at the corner. Then
in equilibrium there is no interaction between the entrant and the incumbents high-quality
product. The incumbent sets pi = q2i and the entrant sets me = qe . Each firm has half of
(I,T T )
(I,T T )
the market. Hence, i
= q4i and e
= 12 qe .
The entrant decides whether to push the incumbent out of the market or co-exist with
it. Comparing the entrants profits under both cases, we have: when qi < 34 qe , the entrant
(I,T T )
(I,T T )
chooses to push out the incumbent, and i
= 0 and e
= qe qi ; otherwise, the
(I,T T )
(I,T T )
entrant chooses to co-exist with the incumbent, and i
= q4i and e
= 12 qe .
Subgame (I, T I): The entrant product will get into Bertrand-type competition with the
xiv
incumbent low-quality free product. In equilibrium, the incumbent is better off not having
the second product as the Bertrand competition eventually kills both incumbents products.
(I,T I)
(I,T I)
= qe qi and i
= 0.
Therefore, in this case me = qe qi . e
Proof of Proposition A-2. Subgame (T, T ): The derivation is straightforward by setting the
FOC of the profit function to zero and solve for the optimal pi and pe . The profits are
4q 2 (qi qe )
(T,T )
(T,T )
i qe (qi qe )
i
= (4qi i q
= q(4q
2 and e
2 .
e)
i qe )
Subgame (T, T T ): We first consider the first case where qi > qi0 > qe . We first take the
FOC of the two profit functions w.r.t. pi , pi0 and pe assuming that qi0 is given. We obtain
(T,T T )
2q (q 0 qe )
q (q q )
4q q q qi 3qe qi0
=
, pi0 = i4q0 0iq
and pe = e4q i00qee . The incumbent profits are i
pi = i i08q 0e2q
e
e
i
qe2 (qi qi0 )+16qi qi20 8qe qi0 (qi +qi0 )

.
4(qi0 qe )2
(T,T T )
w.r.t. qi0 is positive. Hence, it is optimal for

The FOC of i
the incumbent to set qi0 = qi . Effectively, the incumbent will prefer to just offer one product
only. Hence, the optimal tactics are the same as in the (T, T ) case. In this case, (T, T T ) is
weakly dominated by (T, T ).
Subgame (I, T ): The FOC of the incumbent profit function w.r.t. pi gives the optimal price
pi = 21 (qi qe + m2e ). The FOC of the entrant profit function gives the optimal monetization
intensity of product e, me = qi qe . The constraint that qe m2e 0 gives me qe .
Therefore, when qi < 2qe , we have an interior solution. In this case, me = qi qe .

(I,T )
e
= qi q
Substituting it to the expression of equilibrium pi , we have pi = qi qe . Hence, i
2
(I,T )
and e
= 2 qi qe . When qi 2qe , we have a corner solution. In this case, me = qe .
(I,T )
(I,T )
= q4i and e
= 2 qe .
Thus, pi = q2i , i
(I,M )
w.r.t. pi gives pi = 12 (qi qe mi m2i + m2e ).

Subgame (I, M ): The FOC of i
(q q +mi +(m2e m2i ))2
(I,M )
Substituting pi into the profit function, we have: i
= i 4(qei qe +(m
. We can then
2 m2 ))
e
i
take FOC w.r.t. mi and obtain
m2i + qi qe =
m3i
.
mi
(A-1)
The FOC of the entrant profit function w.r.t. me gives:

q
me = qi qe m2i .
(A-2)
We also need qe m2e 0, i.e., me < qe . Hence, when qi 2qe , qi qe m2i qe

and we always have an interior solution. In this case, we could solve equations (A-1) and
xv
(A-2) for mi and me , and obtain the expressions for equilibrium profits.
When qi > 2qe , we may have a corner solution: this happens when me computed from
equation (A-2) is greater than qe . When we are at a corner, me = qe and mi is solved by

equation (A-1).
Subgame (I, I): If qi m2i < qe , the entrant will choose a small me such that qi m2i < qe m2e
and get all the demand. The best response for the incumbent is to decrease mi . Then the
entrant will decrease me . This process ends when qi m2i = qe .
Hence, the equilibrium monetization intensity for the incumbent is mi = qi qe . All
(I,I)
(I,I)
consumers purchase product i. Thus, i
= qi qe and e = 0.
Subgame (I, T T ): When we are in the 2qe < qi case, we know from the (I, T ) case that
the best outcome the incumbent can have is that the incumbent offers one product and the
entrant is at the corner. Hence, the incumbent does not want to offer product i0 .
Now we look at the case where 2qe qi . First consider qi0 > qe . We know that 2qe > qi0 .
Hence, the entrant will be at the interior. From the (I, T ) case, we know that me = qi0 qe .
Now the FOCs of the profit function of the incumbent w.r.t. pi and pi0 give pi = (qe + qi +
m2e )/2 and pi0 = 21 (qe + qi0 + m2e ). The profits of the incumbent are thus 41 (qe + qi + m2e ),
which is independent of qi0 . Hence, the incumbent will set qi0 such that me is as large as
possible. In equilibrium, qi0 = qi and we are back to the case (I, T ).
We then consider the case where qi0 qe . If qi0 < qe ( qi qe )2 , where qi qe is the
equilibrium amount of me without qi0 . Then qi0 has no effect as i0 is killed by the entrant
and the equilibrium outcome is the same as in the (I, T ) case.
Hence, the final case we need to consider is when qe qi0 > qe ( qi qe )2 . In this case,
if the equilibrium me is such that qe m2e > qi0 , i.e., the entrant kills i0 . The case is worse
than the (I, T ) case because the entrant quality is forced to be higher.
Now we look at the case where qe m2e < qi0 . The FOC of the entrant profit function,
p i0
( q 0 (qem
2 ) )me , w.r.t. me is negative. Hence, the entrant wants to lower its monetization
e
i
intensity as much as possible subject to qe m2e < qi0 . Hence, me = qi qe . Essentially,

the entrant kills product i0 . We know that the incumbent prefers not to have product i0 as
having it may force me to be smaller.
To summarize, the incumbent prefers not to offer product i0 and this case is dominated
by the (I, T ) case.
Subgame (I, T I): We know that the two free products will compete as in the (I, I) case and
(I,T I)
the incumbent will push the entrant out of the market. The incumbent maximizes i
by
xvi
setting pi and mi0 . The FOC w.r.t. pi gives: pi = 12 mi0 ( + mi0 ). We then substitute pi into
(I,T I)
i
and obtain:
( + mi0 )2
(I,T I)
.
(A-3)
i
=
4
(I,T I)
It is easy to see that i

increases in mi0 . We conclude that the incumbent will set mi0 to
the maximum. Hence, mi0 = qi qe as the incumbent needs to

make sure that the entrant
(I,T I)
is pushed out. Therefore, the profits are i
2 +(qi qe )+2
4
(qi qe )
(I,T I)
and e
= 0.
Proof of Proposition 1. We first show that (I, I), (I, T I), (I, M ), and (T, T T ) cannot be
equilibrium business model combinations. According to Proposition A-1, we know that
(I, I), (I, T I) and (I, M ) will never the equilibrium outcomes as the entrant will push the
incumbent out of the market in these cases. In addition, when the entrant chooses the
traditional business model, the only possible choices for the incumbent are T and T T and
T T is dominated by T in this case.
According to Proposition A-1, in the case of (T, T ), both firms coexist. In the case of
(I, T ), the entrant kills the incumbent. In the case of (I, T T ), when qe /qi < 4/3, the entrant
chooses to co-exist with the incumbent; otherwise, the entrant chooses to push the incumbent
out and in this case, (I, T ) weakly dominates (I, T T ). Hence, only when the equilibrium
outcome is (I, T ), the incumbent is pushed out of the market.
It is easy see that when is large, the entrant prefers the new model to traditional
model. When the entrant prefers the new model model, the incumbent will be killed when
qe /qi is larger than 4/3 (regardless which business model the incumbent chooses). Hence,
the incumbent will stay put with T and the equilibrium outcome is (I, T ). When the entrant
chooses the new model model and qe /qi is smaller than 4/3, the incumbent can survive only
by choosing T T . Hence, (I, T T ) will be the equilibrium outcome.
Proof of Proposition 2. (I, I) and (I, T I) are never equilibrium outcomes as the entrant will
prefer T to avoid being pushed out of the market. According to Proposition A-2, (I, T T )
and (T, T T ) do not occur in equilibrium because having a second, low-quality product makes
no difference for the incumbents profit.
Part (b) is straightforward. The incumbent cannot be pushed out as it has a higher
quality product and can always price its product close to zero and obtain positive demand.
At the same time, the entrant can always choose to compete through the traditional business
model and obtain some positive profit.
xvii
When qe /qi < 1/2, in the case of (I, M ), the entrant will be at the corner and there will be
no strategic interaction between the entrant and the incumbent. We know as a monopolist,
the incumbent will prefer M to T . In addition, when is not very large, the incumbent
will not want to choose T I or I as the competition between two free products will result in
cannibalization. The entrant will choose I only when is large enough. Hence, (I, M ) is the
outcome when is sufficiently large and qe /qi is small. When qe /qi > 1/2, the incumbent
will prefer T to M as in this case, the entrant is not at the corner and with M , the net
quality levels of incumbent and the entrant will be closer.
Finally, when is large, the incumbent will prefer I if it becomes aware of I and it will
push out the entrant. The entrant prefers not to reveal the innovation and enter with T .
And the best response of the incumbent is then T .
Proof of Proposition 3. Clearly, whenever the entrant adopts the traditional business model,
the competition is the same as in the case without innovation. As the entrant will only choose
to adopt the innovation if it can make more profits than competing with the traditional
model, in all regions where it adopts the innovation, its profitability increases (note that we
assume that when the entrant is indifferent from adopting and not adopting the innovation,
it will choose not to adopt the innovation).
The entrant is at the corner in regions (I, T T ) of Figure 2a and (I, M ) of Figure 2b.
Therefore, the incumbent makes more profits than what it would have earned if the entrant
adopts the traditional business model (as in this case the entrant would not be at the corner).
4qi2 (qi qe )
e
with
innovation
and
In region (I, T ) of Figure 2b, the incumbent profits are qi q
2
(4qi qe )2
without innovation. It is easy to see that the former profits are greater than the latter.
Proof of Proposition 4. In the (I, M ) region where qe < qi , the entrant is at the corner and
the incumbent acts as a monopolist. In this case, the incumbent prefers M to T . Hence, it
earns more profits as a result of the revelation of the innovation by the entrant.
In the case of (I, T T ) where qe > qi , the entrant is at the corner and the incumbent earns
the monopoly profit.
The other two equilibrium outcomes are (T, T ) and (I, T ). In neither case, the entrant
will be at the corner and hence its existence decreases incumbent profits.

Casadesus Masanell, Zhu - Business Model Innovation - R2 PDF

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Casadesus Masanell, Zhu - Business Model Innovation - R2 PDF

Încărcat de

Drepturi de autor:

Formate disponibile

Business Model Innovation and Competitive Imitation: The Case of

Sponsor-Based Business Models

The Harvard community has made this article openly available.

Casadesus-Masanell, Ramon, and Feng Zhu. "Business Model

August 13, 2016 7:34:21 PM EDT

This article was downloaded from Harvard University's DASH

(Article begins on next page)

Business Model Innovation and Competitive Imitation:

October 31, 2011

Business Model Innovation and Competitive Imitation:

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Source: http://www.answers.com/topic/magazine, accessed May 2010.

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

one side imposes a negative externality on the other.

Business Model Innovation and Competitive Imitation

k. Note that the monetization intensity mk is felt as a nuisance by consumersthe greater

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

We think of profit functions as representations of business models as if looked at from a distance. We

Business Model Innovation and Competitive Imitation

subject to 0 1, and qe m2e 0.

the two firms are: i

Business Model Innovation and Competitive Imitation

qe ). In this case, the entrant chooses

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Strategic revelation and strategic concealment

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

LIMITATIONS, EXTENSIONS AND CONCLUSIONS

Business Model Innovation and Competitive Imitation

The figure has qi = 1.

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Source: http://www.reuters.com/finance/stocks/chart?symbol=MTROsdbb.ST, access August

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Gambardella, Alfonso, Anita M. McGahan. 2010. Business-model innovation: General purpose

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Business Model Innovation and Competitive Imitation

Figure 1: The strategic innovation game.

Figure 2: Optimal business models.

Business Model Innovation and Competitive Imitation

(a) Optimal BM without imitation.

(b) Strategic revelation and concealment.